You are here:

Medical Officer-AI Clinical Consultation Roleplay

Jeffrey Cheah School of Medicine and Health Sciences

About this example

In this individual assessment, students will simulate a clinical consultation by interacting with GenAI (ChatGPT) acting strictly as a patient with a specified medical condition. Students select one scenario from a provided list of common and important illnesses, conditions or disorders. During the roleplay, students must:

Conduct a complete and structured medical history
Identify key symptoms and risk factors
Explain the likely biomedical and pathophysiological basis of the condition
Provide appropriate first-line advice or management recommendations within a medical officer’s scope
Demonstrate safe, patient-centred communication throughout

Students will submit:

A transcript of their AI-patient role-play interaction
A written clinical summary consists of:
- Related biomedical, scientific and clinical concepts to the chosen condition
- Explains relevant underlying pathophysiology
- Justifies the treatment or advice provided
- Reflects on the quality and limitations of AI as a simulated patient

Learning outcome:

Relate biomedical, scientific and clinical concepts to common and important illnesses, conditions and disorders and their causes, underlying pathophysiology and treatment.

This assessment was designed to enhance student engagement, give students autonomy in selecting conditions that interest them and provide repeated low-stakes practice before the OSCE exam. GenAI-mediated patient roleplay allows students to practise history taking and apply biomedical knowledge in a safe, flexible and iterative environment. Students can attempt the scenario multiple times and submit their best work.
- Brief students on the assessment expectation and how AI functions as a simulated patient
- Conduct a demonstration using the provided started prompt
- Guide students on ethical and critical use of AI
- Provide guidance on how to write a concise and clinically reasoned summary
- Students feel confident in history taking and clinical communication, as well as reduced anxiety due to the ability to practise repeatedly
- The reflection will ensure students demonstrate biomedical and pathophysiological understanding rather than relying on AI-generated cues

Try it out

This design requires a medium level of effort to implement.

1. Create and test a scenario bank tailored to the conditions taught in the unit
2. Provide a structured AI prompt and safety guidelines
3. Require students to submit both a transcript and reasoning summary (reflection)
4. Use the simulation to prepare students for OSCEs, case-based discussion or viva assessments
- Start by clarifying the learning outcomes and design scenarios that elicit those specific concepts
- Provide clear instructions, structured questioning frameworks and boundaries for AI behaviour
- Encourage iterative practice and gather student feedback to refine the scenarios, prompts and scaffolds
- Ensure the assessment prioritises student reasoning and biomedical explanation, and not the AI performance
Verbal explanation
- For a large class, educators can conduct a peer teaching moment where learners explain principles and primary care to peers allowing educators to observe clarity and accuracy
- For a small class, educators can conduct a 5-10 minutes

Supporting resources

Here are some additional resources that you can browse to help you from TeachHQ:

Learn about AI at your own pace? Enrol in online courses via myDevelopment. Alternatively, you can join ongoing chats about AI and webinars.

Class size: Any	Condition: Individual
Typology: Demonstration	AI statement: Guided use
AI approach: Using and applying AI tools effectively	Visibility of Learning: Verbal explanation

Education Excellence